진동 측정에서의 교정 이해
보정 is the process of comparing a measurement instrument or sensor against a known reference standard of higher accuracy, and documenting the relationship between the instrument’s output and the true value. In 진동 measurement it confirms that an 가속도계, 속도 변환기 또는 분석기 reports the correct value and, where necessary, supplies a correction factor to compensate for any deviation from ideal performance. Calibration is what links a reading on a screen to a traceable physical reality — it underpins quality systems (ISO 9001), legal and contractual compliance, and the integrity of every 상태 모니터링 trend you collect.
Regular calibration matters because 센서 감도 does not stay constant. It drifts with age, temperature cycling, mechanical shock and environmental exposure. An accelerometer that read 100 mV/g when new may, after a hard drop or several years of service, read 96 mV/g — a 4 % error that quietly biases every measurement. Without periodic verification, trend data becomes unreliable, fault-severity calls become inaccurate, and maintenance decisions are made on numbers nobody can defend.
1. Why Calibration Is Necessary
Four distinct needs drive a calibration programme, and a good programme satisfies all of them at once.
- 측정 정확도: sensors drift from their nominal sensitivity — typically 1–5 % per year depending on use — and shock, heat and ageing all accelerate that drift. Verification keeps the reading honest.
- 추적성: an unbroken chain of comparisons links your reading back to a national standard such as NIST (USA) or NPL (UK). The 보정 인증서 documents that chain and is a prerequisite for ISO/IEC 17025 accreditation as well as many legal and contractual obligations.
- 품질 보증: ISO 9001 explicitly requires calibrated measuring equipment. A documented calibration demonstrates that the measurement process is under control and gives confidence in the data used for decisions.
- 일관성: calibrating every sensor to the same reference lets you compare readings from different instruments and trend a machine meaningfully even when the data was gathered with several devices over many years.
2. Calibration Methods
Methods range from absolute laboratory references down to quick functional checks on the shop floor. Each trades accuracy for speed and convenience.
Primary calibration (laser interferometry)
This is the absolute reference method. The sensor is mounted on a precision shaker and its motion is measured directly by a laser interferometer with nanometre resolution; acceleration or velocity is then derived from the measured displacement. It is the most accurate route — uncertainty below 0.5 % — and is performed only by national laboratories and specialised facilities. It is the same interferometric principle exploited by laser vibrometry for non-contact measurement.
Secondary calibration (comparison)
The routine workhorse. The test sensor and a recently primary-calibrated reference sensor are mounted on the same shaker and their outputs compared. Uncertainty is typically 1–3 %, which is more than adequate for the majority of industrial work.
Back-to-back calibration
The test sensor is mounted directly on top of the reference sensor so both experience identical motion, and the two outputs are compared. It is simple, fast and well suited to field verification.
Handheld calibrator
A portable device that generates a precisely known motion — most commonly 1 g at 159.2 Hz (the frequency at which 1 g peak equals 1 mm/s peak velocity, a convenient round number). It is not a full calibration but a rapid confidence check that the sensor and signal chain are alive and reading correctly before critical measurements.
3. The Calibration Certificate
The certificate is the deliverable of any formal calibration and the document an auditor will ask for. A complete 보정 인증서 should record:
- Sensor identification: model and serial number, so the result is tied to a specific physical device.
- 교정 날짜 및 next-due date that defines the validity window.
- 측정된 감도: the actual value (mV/g, pC/g, or mV per mm/s), not the nominal nameplate figure.
- 주파수 응답: the deviation from ideal across the working frequency range.
- Measurement uncertainty: a formal statement of how confident the result is. You can explore how such figures are built up with a measurement uncertainty calculator.
- Traceability and laboratory accreditation: the reference standards used and the lab’s accreditation status.
4. Calibration Intervals and Field Verification
How often to calibrate depends on how much the data matters and how hard the sensor’s life is. Common starting points are: 6–12 months for critical machinery, 1–2 years for general industrial work, 2–3 years for infrequently used instruments, and immediately after any shock or suspected damage. A new sensor’s factory calibration should be verified before it enters service. The interval is then tuned to criticality, usage severity, historical drift rate, environment and any regulatory requirement.
Between formal calibrations, cheap field checks catch gross problems early: a handheld-calibrator check before important work, a back-to-back comparison against a reference sensor, a zero check (output with no input), and consistency checks between sensors reading the same machine. As a rule of thumb, a result within ±2 % of the certificate value is good, within ±5 % is acceptable for most industrial work, and beyond ±10 % calls for recalibration or replacement. A sudden change always warrants investigation — it usually means damage or a connection fault rather than honest drift. To check whether a measured output matches an expected one for a given sensitivity, a 진동 센서 감도 계산기 is a handy companion.
5. Calibration in Practical Field Work
Calibration is not an academic exercise; it is what makes a field reading trustworthy. When an engineer balances a rotor or diagnoses a fault on site, the verdict is only as good as the instrument behind it. A portable two-channel analyser such as the 발란셋-1A ships with sensors of known sensitivity, so the 진폭 및 단계 it reports translate directly into correct correction-weight masses and a defensible pass against the chosen tolerance. Keeping its accelerometers within calibration — and running a quick handheld-calibrator or zero check before a job — is what ensures the residual-vibration figure quoted in a balancing report actually means what it says. The same discipline applies to a 근접 프로브 or any other transducer feeding the analyser.
6. Standards, Records and Best Practice
The governing documents are ISO 16063 (methods for the calibration of vibration and shock transducers), ISO 5347 (accelerometer calibration methods) and ISO/IEC 17025 (general competence of calibration laboratories). Wherever possible use an ISO 17025-accredited lab; accreditation bodies include UKAS in the UK, DKD/DAkkS in Germany and COFRAC in France, with NIST traceability the benchmark in the USA. Accreditation is the practical guarantee that the calibration itself is sound.
Good record-keeping closes the loop. Maintain every certificate, track due dates with automatic reminders, log any out-of-tolerance findings together with the corrective action taken, and trend each sensor’s drift across successive calibrations — a sensor whose sensitivity is creeping in one direction is telling you it will soon need replacing. A centralised calibration database that holds historical data and instrument status makes all of this manageable across a large fleet of sensors.
Finally, treat sensors as the precision instruments they are: protect them from shock and abuse, store them properly, handle cables gently, document any drop, and recalibrate after suspected damage. Calibration is fundamental to measurement quality in vibration analysis — regular comparison against traceable standards, disciplined documentation and systematic field verification are what keep 기준선 and trend data accurate over time, and give the measurement confidence that effective condition monitoring, diagnostics and maintenance decisions ultimately rest on.
